Combustion and Flame 157 (2010) 2137–2154

Contents lists available at ScienceDirect

Combustion and Flame

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / c o m b u s t fl a m e

An experimental and kinetic modeling study of combustion of isomers of butanol

Roberto Grana a, Alessio Frassoldati a, Tiziano Faravelli a, Ulrich Niemann b, Eliseo Ranzi a,*, Reinhard Seiser b,
Robert Cattolica b, Kalyanasundaram Seshadri b
a
Dipartimento di Chimica, Materiali e Ingegneria Chimica, Politecnico di Milano, Milano, Italy
b
Department of Mechanical and Aerospace Engineering, University of California at San Diego, La Jolla, CA 92093-0411, USA

a r t i c l e i n f o a b s t r a c t

Article history: A kinetic model is developed to describe combustion of isomers of butanol—n-butanol (n-C4H9OH), sec-
Received 16 February 2010 butanol (sec-C4H9OH), iso-butanol (iso-C4H9OH), and tert-butanol (tert-C4H9OH). A hierarchical approach
Received in revised form 22 April 2010 is employed here. This approach was previously found to be useful for developing detailed and semi-
Accepted 17 May 2010
detailed mechanism of oxidation of various hydrocarbon fuels. This method starts from lower molecular
Available online 10 June 2010
weight compounds of a family of species and proceeds to higher molecular weight compounds. The pyro-
lysis and oxidation mechanisms of butanol isomers are similar to those for hydrocarbon fuels. Here, the
Keywords:
development of the complete set of the primary propagation reactions for butanol isomers proceeds from
Butanol isomers
Kinetic modeling
the extension of the kinetic parameters for similar reactions already studied and recently revised for eth-
Flame structure anol, n-propanol and iso-propanol. A detailed description leading to evaluation of rate constants for ini-
Bio-fuels tiation reactions, metathesis reactions, decomposition reactions of alkoxy radicals, isomerization
reactions, and four-center molecular dehydration reactions are given. Decomposition and oxidation of
primary intermediate products are described using a previously developed semi-detailed kinetic model
for hydrocarbon fuels. The kinetic mechanism is made up of more than 7000 reactions among 300 spe-
cies. The model is validated by comparing predictions made using this kinetic model with previous and
new experimental data on counterflow non-premixed flames of n-butanol and iso-butanol. The struc-
tures of these flames were measured by removing gas samples from the flame and analyzing them using
a gas chromatograph. Temperature profiles were measured using coated thermocouples. The flame struc-
tures were measured under similar conditions for both fuels to elucidate the similarities and differences
in combustion characteristics of the two isomers. The profiles measured include those of butanol, oxygen,
carbon dioxide, water vapor, carbon monoxide, hydrogen, formaldehyde, acetaldehyde, and a number of
C1–C4 hydrocarbon compounds. The predictions of the kinetic model of flame structures of the two iso-
mers were satisfactory. Validation of the kinetic model was also performed by comparing predictions
with experimental data reported in the literature. These data were obtained in batch reactors, flow reac-
tors, jet-stirred reactors, and shock tubes. In these configurations, combustion is not influenced by molec-
ular transport. The agreement between the kinetic model and experimental data was satisfactory.
Ó 2010 The Combustion Institute. Published by Elsevier Inc. All rights reserved.

1. Introduction use of fossil fuels. Recent legislation in many countries require

Recent interest in promoting the use of bio-fuels arises from the
need to improve energy security and reduce net greenhouse gas
emissions. At the present rate of energy consumption, worldwide
reserves of natural gas, oils and to a limited extent coal are rapidly
diminishing. This process is being accelerated by significant in-
creases in the rates of energy consumption in developing countries.
These developments have increased the cost of fossil fuels and
have an adverse impact on the national economies of the world.
Thus, there is a need to decrease the use of fossil fuels for sustain-
able development, and allow future generations to continue the
* Corresponding author.
E-mail address: eliseo.ranzi@polimi.it (E. Ranzi).
them to limit emissions of greenhouse gases. Bio-fuels are carbon
neutral, thus they are considered to be more environmentally
friendly and help in meeting legislative requirements by limiting
emissions of greenhouse gases. Both EU and USA are committed
to reducing energy consumption and to increase renewable fuel
use. All these point to increasing use of bio-fuels in the near future.
Alcohols show significant potential to be an alternative to con-
ventional gasoline. Alcohols are renewable fuels because they can
be produced from biomass fermentation and are by-products of
Fischer Tropsch processes. Ethanol is currently a component of
reformulated gasoline. The amount of ethanol in gasoline is pro-
jected to increase in the future. There is considerable interest in
promoting the use of butanol (C4H9OH) as an alternative to etha-
nol. Butanol can be derived from lignocellulosic materials. Butanol

0010-2180/$ - see front matter Ó 2010 The Combustion Institute. Published by Elsevier Inc. All rights reserved.
doi:10.1016/j.combustflame.2010.05.009
2138 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154
has some advantages, as transportation fuel component, when
compared with ethanol. It is less corrosive, has a lower vapor
pressure, higher energy density, and its octane rating is similar
to that of gasoline. Thus, it can be blended with gasoline at much
higher proportions than ethanol without compromising efficiency.
Butanol is immiscible when mixed with water at concentrations
higher than about 7–8%. This eliminates a number of storage con-
cerns and makes phase separation considerably easier in compar-
ison to ethanol. Unfortunately, butanol has a foul odor that can
persist for a long time. Butanol is yet to be employed as exten-
sively as ethanol as transportation fuel, because the production
of butanol is labor intensive and only low yields have been
achieved. These limitations are expected to improve in the near
future.
Numerous studies have addressed combustion of methanol
[1–6], ethanol [1,7–14] and propanol isomers [15]. These studies
provide the building blocks for the kinetic modeling of combustion
of butanol isomers. The isomers of butanol are n-butanol (n-
C4H9OH), sec-butanol (sec-C4H9OH), iso-butanol (iso-C4H9OH),
and tert-butanol (tert-C4H9OH). Dagaut and Togbé [16] developed
a sub-mechanism for n-butanol, which was combined with an
overall kinetic mechanism of oxidation of gasoline. Moss et al.
[17] have proposed a detailed kinetic scheme for describing the
high-temperature oxidation of the different butanol isomers. Re-
cently, Black et al. [18] have tested predictions of a kinetic model
of n-butanol oxidation with autoignition delay times measured in
shock tubes, and data on evolution of various species measured
in a jet-stirred reactor, with encouraging results. These studies
[16–18] were restricted to premixed systems without flow. Studies
on premixed combustion with molecular transport are available
[19,20]. McEnally and Pfefferle [19] studied pollutant emissions
from methane/air flames doped with the four butanol isomers.
Yang et al. [20] studied laminar premixed, low-pressure flames of
the four butanol isomers, by using photoionization mass spectrom-
etry. The structures of non-premixed n-butanol flames in non-uni-
form flows were measured previously employing the counterflow
configuration [21]. The structure of non-premixed flames depends
on the stoichiometric mixture fraction, Zst, and the strain rate [22–
24]. The previous measurements of the structure of n-butanol
flames were made with the oxidizer stream made up of air en-
riched with oxygen [21]. The experimental conditions were charac-
terized by high values of the stoichiometric mixture fraction and
low values of the strain rates. Combustion in practical systems
are characterized by low values of the stoichiometric mixture frac-
tion and both low and high values of the strain rates [25]. To com-
plement the previous studies, new experimental data on the
structures of counterflow non-premixed flames of n-butanol and
iso-butanol are shown here. The flame structures are measured
with air as the oxidizer. The experimental conditions are character-
ized by low values of stoichiometric mixture fraction, and moder-
ately high values of strain rate. The new experimental data allows
comparison of the flame structures of butanol isomers. The new
experimental data together with previous experimental data are
helpful for validating the kinetic model of alcohol fuels.
A hierarchical approach is found to be useful for developing de-
tailed and semi-detailed mechanism for describing oxidation of
various fuels. This method starts from lower molecular weight
compounds of a family of species and proceeds to higher molecular
weight compounds. This procedure allows extensions to other
compounds using similarity and analogy rules. Using this proce-
dure, validated kinetic mechanisms of oxidation of ethanol [14]
and propanol isomers have been developed [15]. Here, an oxida-
tion scheme that considers the primary reactions of the four buta-
nol isomers is proposed and tested. This kinetic mechanism is
analogue and aligned with those of the lower alcohols. This is a
further step towards the confirmation of the proposed classes of
reactions and of their rate constants, with a particular attention
to H abstraction and molecular dehydration reactions.
2. Experimental measurements of flame structure
The structures of non-premixed flames of n-butanol and iso-
butanol were measured employing the counterflow configuration.
Fig. 1 shows a schematic illustration of the counterflow configura-
tion. Steady, axisymmetric, laminar flow of two counterflowing
streams toward a stagnation plane is considered. In this configura-
tion, a fuel stream made up of prevaporized fuel (n-butanol or iso-
butanol) and nitrogen is injected from the fuel duct, and an oxi-
dizer stream of air is injected from the oxidizer duct. These jets
flow into the mixing layer between the two ducts. The exit of the
fuel duct is called the fuel boundary and the exit of the oxidizer
duct the oxidizer boundary. Fine wire meshes are placed at the ex-
its of the ducts. As a consequence, the tangential component of the
flow velocities vanishes at the boundaries. This allows the use of
‘‘plug flow boundary conditions” in the numerical simulations of
the flame structure. The mass fraction of fuel, the temperature,
and the component of the flow velocity normal to the stagnation
plane at the fuel boundary are represented by YF,1, T1, and V1,
respectively. The mass fraction of oxygen, the temperature, and
the component of the flow velocity normal to the stagnation plane
at the oxidizer boundary are represented by YO2,2, T2, and V2,
respectively. The exit diameter of the fuel duct and the oxidizer
duct is 23.3 mm. The distance between the fuel boundary and
the oxidizer boundary is represented by L.
The value of the strain rate, defined as the normal gradient of
the normal component of the flow velocity, changes from the fuel
boundary to the oxidizer boundary [26]. The characteristic strain
rate on the oxidizer side of the stagnation plane a2 is given by [26]
pffiffiffiffiffiffi pffiffiffiffiffiffi
a2 ¼ ð2jV 2 j=LÞb1 þ jV 1 j q1 =ðjV 2 j q2 Þc: ð1Þ
Here, q1 and q2 represent the density of the mixture at the fuel
boundary and at the oxidizer boundary, respectively. The stoichi-
ometric mixture fraction, Zst is [22–24]
Z st ¼ ð1 þ mY F;1 =Y O2 ;2 Þ1 : ð2Þ
where m is the stoichiometric mass ratio of oxygen to fuel. For stoi-
chiometric combustion of butanol, m = 2.5946.
The profiles of concentration of stable species were measured
for YF,1 = 0.3, T1 = 353 K, YO2,2 = 0.233, T2 = 298 K, a2 = 100 s1,
V1 = 0.248 m/s, V2 = 0.25 m/s, and L = 10 mm. At these conditions,
the flame is on the oxidizer side of the stagnation plane with the
stoichiometric mixture fraction Zst = 0.23. Fig. 2 shows the photo-
graph of a n-butanol flame stabilized in the counterflow burner.
Concentrations profiles of stable species were measured by remov-
Fig. 1. Schematic illustration of the counterflow configuration. The figure shows the
thermocouple, used for measuring flame temperature.
 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154
Fig. 2. Photograph of a non-premixed n-butanol flame stabilized in the counterflow
burner for YF,1 = 0.3, T1 = 353 K, YO2,2 = 0.233, T2 = 298 K, a2 = 100 s1, V1 = 0.248 m/s,
V2 = 0.25 m/s, and L = 10 mm. The figure shows the quartz microprobe.
ing gas samples from the reaction zone using a heated quartz
microprobe, and analyzing them in a gas chromatograph. All lines
from the sample probe to the gas chromatograph were heated to
373 K. The microprobe has a tip with an inner diameter of
150 lm. To minimize disturbances to the flow-field, the tip of
the microprobe was placed at a location of 5 mm off the axis of
symmetry as shown in Figs. 1 and 2. The location of the sampling
probe in the flow-field was determined using a digital photo
camera.
The size of one pixel in the camera corresponds to a distance of
approximately 20 lm in the flow-field. The mole fractions of vari-
ous species in the sample were measured using an Agilent 3000mi-
croGC gas chromatograph. This instrument is equipped with a
10 m long molecular sieve 5A column, a 8 m long Poraplot U col-
umn, a 8 m long Poraplot Q column, and a 8 m long OV-1 column.
The gas chromatograph has a built-in sample pump. To ensure
equal sample sizes and thus comparable results across all measure-
ments, a constant sample inlet pressure is required. Therefore, spe-
cies were sampled at a constant pressure of 600 mbar into a
sample vessel. Nitrogen was then introduced until the vessel at-
tained a total pressure of 1150 mbar. After a waiting period of
6 min, to allow sufficient mixing, the sample was introduced into
the gas chromatograph. The molecular sieve column uses argon
as a carrier gas. It was used to separate hydrogen (H2), oxygen
(O2), nitrogen (N2), methane (CH4), and carbon monoxide (CO).
All other columns use helium as a carrier gas. The Poraplot U
column was used for separating carbon dioxide (CO2), ethene
(C2H4), ethane (C2H6), ethyne (C2H2), and formaldehyde (CH2O).
The Poraplot Q column was used for separating water (H2O),
1,2-propadiene (C3H4), propyne (C3H4), propene (C3H6), propane
(C3H8), acetaldehyde (C2H4O), butene (C4H8), and butadiene
(C4H6). The OV-1 column was used for separating butanal
(n-C3H7CHO), n-butanol and iso-butanol. The mole fractions of var-
ious species eluting from the columns were measured using ther-
mal-conductivity detectors (TCD). The detectors were calibrated
using samples of known composition. For those species that are
gases at 298 K and pressure of 1 bar, the calibration was performed
with calibration gases of known composition. For those species
that are liquids at 298 K and 1 bar calibration was performed using
a sample vessel. First, the sample vessel was evacuated. Next, the
liquid was injected into the vessel with a syringe through a septum
placed in the wall of the vessel. It was then diluted with nitrogen.
The concentration of the species was established from the pressure
of the vessel recorded after evaporation and the total pressure after
nitrogen dilution. This procedure was used for n-butanol, iso-buta-
nol, water, and formaldehyde. The peaks for all species presented
here show very good separation. Therefore the expected accuracy
for the maximum concentrations of all species except H2O and
2139
CH2O is expected to be better than ±10%. The expected accuracy
for H2O and CH2O is ±20%. The accuracy for formaldehyde is based
on the signal size compared to the baseline noise. The calibration
for formaldehyde showed very little deviation and good
repeatability.
Temperature profiles were measured for n-butanol and
iso-butanol flames, using a Platinum–Platinum13%Rhodium
thermocouple (R-type), at conditions identical to those employed
in the measurement of concentration profiles. As shown in Fig. 1,
the measurements were made along a line that is parallel to the
axis of symmetry and approximately 5 mm away from the axis.
The exact location of the thermocouple bead was determined using
a digital photo camera. The measurements were made with bare
thermocouple wires (uncoated) as well as with coated thermocou-
ple wires. The wire diameter of the bare (uncoated) thermocouple
was 25 lm and the bead diameter was 90 lm. The coated thermo-
couple had a layer of BeO/Y2O3 to avoid catalytic reactions at the
surface of the thermocouple [27]. The coating was performed fol-
lowing the procedure recommended in Ref. [27]. The wire diame-
ter of the coated thermocouple was 35 lm and the bead
diameter was 135 lm. The measured temperatures with both ther-
mocouples were corrected by taking into consideration radiative
heat losses from the surface of the thermocouple. The corrections
were made employing the procedure described by Peterson and
Laurendeau [28]. The convective heat transfer from the gas to the
thermocouple was estimated assuming that the thermocouple is
a cylinder placed in cross flow, with the Nusselt number of 0.5.
The Reynolds number used for estimating the Nusselt number
was evaluated using the wire diameters as the characteristic
length, the characteristic velocity was V2 and the kinematic viscos-
ity was estimated at a temperature of 1500 K. The Reynolds num-
ber was 0.0055 for the uncoated thermocouple and 0.0075 for the
coated thermocouple. The Prandtl number was 0.7. The heat losses
from the wire by radiation were estimated assuming an emissivity
of 0.2 for the uncoated wire and 0.6 for the coated wire. At the peak
value of the measured temperature, the radiation correction was
approximately 50 K for bare thermocouple and 100 K for the
coated thermocouple. The absolute accuracy of the temperature
measurement is expected to be better than ±80 K. The flame struc-
ture of n-butanol is shown in Figs. 20 and 21 and that of iso-buta-
nol in Figs. 23 and 24. The comparisons with predictions of the
kinetic model are discussed later.
The experimental data obtained here are compared with the
experimental data of Sarathy et al. [21]. The exit diameter of the
fuel duct and the oxidizer duct of the counterflow burner em-
ployed in this previous study was 25.4 mm, and the separation dis-
tance L = 20 mm. The fuel stream was a mixture of 94.11% N2 and
5.89% n-butanol by volume (YF,1 = 0.142) with T1 = 356 K. The mass
flow rate of the fuel stream was 0.131 kg/(m2 s). The oxidizer
stream was a mixture of 42.25% O2 and 57.75% N2 by volume
(YO2,2 = 0.4554) with T2 = 423 K. The mass flow rate of the oxidizer
stream was 0.126 kg/(m2 s). At these conditions, the flame is on the
fuel side of the stagnation plane with the stoichiometric mixture
fraction Zst = 0.5528, and the strain rate at the stagnation plane
in the region between the stagnation plane and the fuel boundary
was a1 = a2(q2/q1)0.5 = 33 s1. The flame structure of this n-butanol
flame [21] is shown in Fig. 22.
3. Kinetic mechanism and reaction classes
Figs. 3–6 show simplified primary decomposition mechanisms
for the four isomers of butanol—n-butanol, sec-butanol, iso-buta-
nol, and tert-butanol. As shown in Fig. 3, five different radicals
are formed via H-abstraction reactions from n-C4H9OH. Subse-
quently, these radicals isomerise and decompose. 1-Butene
2140 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154

-H2O -H2
O
OH

-H -H -H -H -H
-CH2CH2OH

O OH OH OH OH

-C3H7 -H -H -C2H5 -H -H -OH -CH3 -H -H -CH2OH -H -H

CH2O O OH OH OH OH
OH

Fig. 3. Primary decomposition reactions of n-butanol.

O
OH -H2
-H2O

-H
-OH
-H -H -H -H -CH3CHOH

OH OH O OH OH

-C2H5 -H -H -CH3 -H -H -H -CH3 -C2H5 -OH -H -CH3 -H -H

O OH OH OH
OH OH OH O O

OH

Fig. 4. Primary decomposition reactions of sec-butanol.

-H2O -H2
OH O

-H -H -H -H

O OH OH OH

-C3H7 -H -H -CH3 -H -H -OH -H -H -CH2OH -CH3

CH O O
OH OH OH OH
2

Fig. 5. Primary decomposition reactions of iso-butanol.

 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154 2141

Pressure dependence is not included in these expressions. How-

-H2O
OH ever, it is worth having a caution, as high temperatures or very
low pressures are considered.

3.1. Unimolecular initiation reactions

-H -H -OH

The activation energy of initiation reactions are evaluated from

O OH the strength of the C–C bond by defining the bond energy of pri-
mary, secondary and tertiary C atoms (Cp, Cs, and Ct) with the dif-
ferent C atoms with OH substitution (C(OH)p/s/t). The four butanol
isomers are good examples to define and illustrate this reaction
-CH3 -CH3 class of unimolecular initiation reactions:
n-C4 H9 OHn-C3 H7 þ  CH2 OH k ¼ 2:0  1016 exp½85; 000=ðRTÞ½s1 ;
O OH
sec-C4 H9 OHCH3 þ C2 H5 C HOH k ¼ 2:0  1016 exp½87; 000=ðRTÞ½s1 ;
sec-C4 H9 OHC2 H5 þ CH3 C HOH k ¼ 2:0  1016 exp½86; 500=ðRTÞ½s1 ;
iso-C4 H9 OHiso-C3 H7 þ  CH2 OH k ¼ 2:0  1016 exp½85; 000=ðRTÞ½s1 ;
Fig. 6. Primary decomposition reactions of tert-butanol.
tert-C4 H9 OHCH3 þ ðCH3 Þ2C OH k ¼ 6:0  1016 exp½81; 500=ðRTÞ½s1 :

(1-C4H8) is the result of the molecular dehydration reaction as well Kinetic data for the remaining initiation reactions, that involve the
as the dehydroxylation reaction of n-C4H8OHb radical. Butanal splitting of the C–C bond, are similar to those used for alkanes. The
(n-C3H7CHO) and/or butenyl alcohols are formed from dehydroge- following kinetic parameters are prescribed for the reactions that
nation reactions of all the five primary radicals. Butanal, methyl- result in formation of the OH radical:
ethyl-ketone (MEK) and methylpropanal are also formed by molec-
ular dehydrogenation of n-butanol, sec-butanol, and iso-butanol, n-C4 H9 OHn-C4 H9 þ  OH k ¼ 1:5  1016 exp½93; 200=ðRTÞ½s1 ;
respectively. Allyl and vinyl alcohols are formed from b-decompo- sec-C4 H9 OHsec-C4 H9 þ  OH k ¼ 1:5  1016 exp½93; 200=ðRTÞ½s1 ;
sition reactions of n-C4H8OHa and n-C4H8OHb. a-Unsaturated iso-C4 H9 OHiso-C4 H9 þ  OH k ¼ 1:5  1016 exp½93; 200=ðRTÞ½s1 ;
alcohols, usually formed at high temperatures, are considered as
tert-C4 H9 OHtert-C4 H9 þ  OH k ¼ 1:5  1016 exp½93; 200=ðRTÞ½s1 :
directly transformed into the corresponding aldehydes, through
the keto-enol tautomerization, which is quite fast in these condi-
They are the same as that for the similar propanol initiation reactions.
tions. This allows to reduce the total number of species in the over-
Note that these reference kinetic parameters are not affected by the
all kinetic scheme.
nature of the C atom. High activation energies are required to release
For the sake of brevity, isomerisation reactions among the pri-
the H atoms from the OH group or the carbon skeleton, therefore reac-
mary butanol radicals are not reported in this scheme. Similarly,
tions of this type cannot contribute to fuel decomposition. On the
as shown in Fig. 4, sec-C4H9OH forms five different primary radi-
contrary, the reverse reactions could affect flame propagation and
cals and MEK is the fingerprint of this alcohol. Fig. 5 shows that
all these reactions are included in the overall mechanism with the
there are four primary radicals of iso-butanol, while only two rad-
same kinetic parameter k = 5.0  1010 [l mol1 s1].
icals are formed from tert-butanol (Fig. 6).
The pyrolysis and oxidation mechanisms of butanol isomers are
3.2. Metathesis reactions
similar to those for hydrocarbon fuels. Here the development of the
complete set of the primary propagation reactions for butanol iso-
Metathesis reactions are treated according to the systematic ap-
mers proceeds from the extension of the kinetic parameters for
proach described elsewhere [29]. Following the kinetic model for
similar reactions already studied and recently revised for ethanol,
propanol isomers [15], the kinetic parameters for the H-atom
n-propanol and iso-propanol [1,7–15]. The kinetic study of oxida-
abstraction from the alcohol functional group are assumed to be
tion of n-propanol and iso-propanol is a useful starting point for
equal to those for the abstraction of a primary H atom from a
the extension of the kinetic scheme to butanol isomers [15]. Initi-
methyl group. Galano et al. [30] studied the gas phase reactions
ation reactions are, in general, evaluated by assuming a reference
of alcohols with the OH radical employing a quantum mechanical
frequency factor, A, with the activation energy, E equal to the bond
approach. Different from the previous recommended values
energy, and microscopic reversibility based on the reverse radical
[31,32], they concluded that the rate coefficient corresponding to
recombination reaction is applied. Metathesis reactions require
the a channel (ka) is larger than those of the other competing
defining the reactivity of the H atoms in hydroxyl position and
channels. Further considerations of Galano et al. [30] relating to
the H atoms in a position. Remaining H atoms are presumed to
the higher reactivity of the c-site at low temperatures could be
be unaffected by the presence of the OH group. Decomposition
simply attributed to the higher reactivity of the (x  1) sites and
reactions of the corresponding alkoxy and parent radicals from
are neglected here. Similar arguments were also recently put for-
alcohol fuels need further elaboration. Isomerization reactions of
ward by Black et al. [18]. On the basis of these previous studies,
these radicals are significant and could enhance the role of the very
we assume the following kinetic parameters for the H-abstraction
reactive alkoxy radical. Finally, the class of the four-center molec-
reactions of OH radicals in a position (units are: cal, l, mole, s, K):
ular dehydration reactions requires a careful and systematic dis-
cussion in order to highlight the role of primary, secondary and 
OH þ n-C4 H9 OHH2 O þ n-C4 H8 OHa
tertiary sites. In this section, unimolecular reactions are discussed 1
k ¼ 1200  T 2  exp½2260=ðRTÞ½l mol s1 ;
first followed by metatheses reactions, decomposition and isomer-
ization reactions of primary radicals from alcohol fuels, and four-

OH þ sec-C4 H9 OHH2 O þ sec-C4 H8 OHa
1
center molecular dehydration reactions. The rate constant, ki for k ¼ 600  T 2  exp½3180=ðRTÞ½l mol s1 ;
reaction i is written as ki = AiTni exp[Ei/(RT)], where Ai is the fre- 
OH þ iso-C4 H9 OHH2 O þ iso-C4 H8 OHa
quency factor, Ei is the activation energy in cal/mol, T the temper- 1
ature, ni the temperature exponent, and R is the gas constant. k ¼ 1200  T 2  exp½2260=ðRTÞ½l mol s1 :
2142 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154
Thus, kinetic parameters of H abstractions in a position from
n-butanol and iso-butanol are obtained by increasing by 50% the
frequency factors of the two secondary H atoms, while the param-
eters for sec-butanol are obtained by increasing by 50% the
frequency factor of the tertiary H atom. As an example, the selectiv-
ity of n-butanol radicals is n-C4H8OHa > n-C4H8OHb  n-
C4H8OHc > n-C4H8OHd > n-C4H9O.
For the other H sites, the systematic approach for metathesis
reactions, described elsewhere is maintained [29,33]. Metathesis
reactions of H, CH3 and other abstracting radicals are treated fol-
lowing the same approach.
3.3. Decomposition reactions of alkoxy radicals
The kinetic parameters of alkoxy radicals decomposition reac-
tions via b-scission have been recently discussed by Curran [34]
and Frassoldati et al. [15]. Alkoxy radical decompositions to form
alkyl radicals prevail over the dehydrogenation reaction paths.
The analogy with propanol decomposition reactions suggests the
use of the following kinetic parameters:
n-C4 H9 O n-C3 H7 þ CH2 O k ¼ 3:0  1013 exp½15; 000=ðRTÞ½s1 
n-C4 H9 O H þ C3 H7 CHO k ¼ 3:0  1013 exp½26; 000=ðRTÞ½s1 
sec-C4 H9 O H þ CH3 COC2 H5 k ¼ 3:0  1013 exp½23; 000=ðRTÞ½s1 
sec-C4 H9 O CH3 þ C2 H5 CHO k ¼ 3:0  1013 exp½18; 000=ðRTÞ½s1 
sec-C4 H9 O C2 H5 þ CH3 CHO k ¼ 3:0  1013 exp½15; 000=ðRTÞ½s1 
iso-C4 H9 O iso-C3 H7 þ CH2 O k ¼ 3:0  1013 exp½20; 000=ðRTÞ½s1 
  13
iso-C4 H9 O H þ ðCH3 Þ2 CHCHO k ¼ 3:0  10 exp½26; 000=ðRTÞ½s 
tert-C4 H9 O CH3 þ CH3 COCH3 k ¼ 9:0  1013 exp½18; 000=ðRTÞ½s1 :
These rate expressions highlight the high reactivity of alcoxy radi-
cals and the low selectivity toward the dehydrogenation channel.
In the temperature range of 700–1000 K decomposition reactions
are at least 10 times faster. Kinetic data for other decomposition
and dehydrogenation reactions of alkoxy and alkyl-hydroxy radicals
are the same as those used for propanol and alkane fuels.
3.4. Isomerization reactions
Isomerization reactions of primary radicals can play an impor-
tant role during the first decomposition steps, mainly due to the
high reactivity of butoxy radicals. In fact, 1–4 and 1–5 H transfer
reactions are explained on the basis of internal H-abstraction reac-
tions, via 5-membered and 6-membered ring intermediates. The
rate constants of these isomerization reactions are estimated in
terms of the number of atoms in the transition state ring structure
(including the H atom) and the type of sites involved in the H
transfer [35,36]. Based on similar reactions of alkyl radicals, the fol-
lowing isomerization reactions of n-C4H8OHd are considered:
n-C4 H8 OHdn-C4 H9 O k1;5 ¼ 1:6  1010 exp½16; 000=ðRTÞ½s1 ;
n-C4 H8 OHdn-C4 H8 OHa k1;4 ¼ 2:0  1011 exp½19; 600=ðRTÞ½s1 :
Values for the rate parameters for the isomerization reactions are
obtained by using the reference kinetic parameters, accounting for
the extra activation energy for the ring strain and decreasing the
frequency factor for the tie up of the additional rotors [37]. Fig. 7
compares the values of rate parameters of these isomerization reac-
tions with the b-decomposition and the dehydrogenation reaction
of n-C4H8OHd:
n-C4 H8 OHd CH2 CH2 OHþC2 H4 kdec ¼3:01013 exp½30;000=ðRTÞ½s1 ;
 13
n-C4 H8 OHdH þC3 H5 CH2 OHkdehyd ¼3:010 exp½36;000=ðRTÞ½s :
10 8
β-decomposition
Rate constant [s-1 ]
10 7
1,5-isomerization
1,4-isomerization dehydrogenation
10 6
800 900 1000 1100 1200
Temperature [K]
Fig. 7. Isomerization and decomposition reactions of n-C4H8OHd radical.
Decomposition reactions prevail only at temperatures higher than
1000–1100 K, while the 1–5 isomerization reaction dominates at
temperatures lower than 900 K. Moreover, dehydrogenation reac-
tion could compete only at flame temperatures. Of course, these
isomerization reactions are more effective when less reactive radi-
cals form the unstable alkoxy radicals.
3.5. Four-center molecular dehydration reactions
This class of reactions involves a four-center cyclic transition
1 state with the formation of parent alkenes and H2O, as shown in
Fig. 8. Thus sec-butanol dehydration can form either 1-butene or
2-butene, depending on the different four membered ring transi-
tion intermediate. Several kinetic parameters have been suggested
for this class of reactions [9,10,16,17,38–40]. Different activation
energies and frequency factors are expected depending on the type
of OH and the number and the type of H involved.
Moss et al. [17] have proposed different values for the different
butanol isomers. They used for iso-butanol the kinetic parameters
proposed by Bui et al. [40] for iso-propanol, while the rate parame-
ters for the other isomers were fitted in order to obtain sufficient
agreement with their experimental ignition delay times. They also
observed a disagreement between all the fitted rate constants and
suggested the need for a better analysis of these reactions, mainly
due to their large sensitivity. In order to define a general rule for this
class of reactions, it is first useful to select the kinetic parameters of a
reference reaction, where the dehydration involves a primary OH
group and a single primary H atom. On the basis of the previous ki-
netic studies, these reference kinetic parameters are assumed
k ¼ 5:0  1013 exp½68; 600=ðRTÞ½s1 
The frequency factor will account for the number of H atoms while
the activation energy reflects the type of H atoms and the type of
OH group. A reduction of activation energy of 1000 and 3000 cal/
mol is assumed for secondary and tertiary H atoms, respectively.
OH
H
+ H2O
OH
OH
H
+ H2O
1
Fig. 8. Dehydration reactions of sec-butanol to form 1-butene and 2-butene.
 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154 2143

These corrections reveal the differences in the bond energy of the
released chemical bonds. According to Markovnikov’s rule and the
kinetic study of Dente [41] the corrections relating the type of the
OH group reflect the different ionization propensity of the substi-
tuted C atoms. With the addition of a protic acid such as H–X to
an alkene, the acid hydrogen (H) attaches to the C atom with the
greatest number of hydrogen atoms, and the halide group (X) at-
taches to the more substituted C atom. Due to the analogy of H–
OH additions and the microscopic reversibility of elementary reac-
tions, corrections of 1500 and 3500 kcal/kmol are assumed when
secondary and tertiary OH groups are involved in the cyclic transi-
tion state. Thus, the following kinetic parameters of the dehydration
reactions are assumed for the different butanol isomers:
n-C4 H9 OHn-C4 H8 þ H2 O k ¼ 1:0  1014 exp½67; 600=ðRTÞ½s1 ;
sec-C4 H9 OH2-C4 H8 þ H2 O k ¼ 1:0  1014 exp½66; 100=ðRTÞ½s1 ;
sec-C4 H9 OH1-C4 H8 þ H2 O k ¼ 1:5  1014 exp½67; 100=ðRTÞ½s1 ;
iso-C4 H9 OHiso-C4 H8 þ H2 O k ¼ 5:0  1013 exp½65; 600=ðRTÞ½s1 ;
tert-C4 H9 OHiso-C4 H8 þ H2 O k ¼ 4:5  1014 exp½65; 100=ðRTÞ½s1 :
Dehydrogenation reactions are another class of four-center molecu-
lar reactions. These and similar reactions have been extensively and
systematically studied by Dente [41]. Dehydrogenation reactions
directly form acyl groups and could become of interest mainly at
low pressures. As an example, the following molecular dehydroge-
nation reactions are considered:
n-C4 H9 OHC3 H7 CHO þ H2 k ¼ 5:0  1013 exp½69; 500=ðRTÞ½s1 ;
sec-C4 H9 OHCH3 COC2 H5 þ H2 k ¼ 5:0  1013 exp½69; 500=ðRTÞ½s1 ;
iso-C4 H9 OHðCH3 Þ2 CHCHO þ H2 k ¼ 5:0  1013 exp½69; 500=ðRTÞ½s1 :
Due to the formation of acyl group, these reactions require lower
activation energy, when compared to the corresponding ones of al-
kanes [42].
n-C4 H10 n-C4 H8 þ H2 k ¼ 5:0  1013 exp½71; 000=ðRTÞ½s1 :
The detailed sub-mechanism of the four butanol isomers is reported
in Table 1, together with the kinetic parameters. Further pyrolysis
and/or oxidation reactions of intermediate products are described
in a semi-detailed oxidation mechanism for hydrocarbon fuels up
to C16 developed previously [43–45]. The overall kinetic scheme is
based on hierarchical modularity and is made up of more than
7000 reactions among 300 species. Thermochemical data for most
species was obtained from the CHEMKIN thermodynamic database
[46,47]. For those species for which thermodynamic data is not
available in the literature, the group additive method was used to
estimate these properties [35]. The complete mechanism, with

Table 1
Sub-mechanism of butanol isomers [units are: l, mol, s, cal].

Reactions A n E
n-Butanol
nC4 H9 OH ¡ nC3 H7 þ  CH2 OH 2.0  10+16 0 85,000
nC4 H9 OH ¡ C2 H5 þ  CH2 CH2 OH 2.0  10+16 0 85,000
nC4 H9 OH ¡ CH3 þ  CH2 CH2 CH2 OH 2.0  10+16 0 86,000
nC4 H9 OH ¡  OH þ  CH2 CH2 CH2 CH3 1.5  10+16 0 93,200
nC4 H8 OHa þ H ¡ nC4 H9 OH 5.0  10+10 0 0
nC4 H8 OHb þ H ¡ nC4 H9 OH 5.0  10+10 0 0
nC4 H8 OHc þ H ¡ nC4 H9 OH 5.0  10+10 0 0
nC4 H8 OHd þ H ¡ nC4 H9 OH 5.0  10+10 0 0
nC4 H9 O þ H ¡ nC4 H9 OH 5.0  10+10 0 0
nC4 H9 OH ¡ H2 O þ nC4 H8 1.0  10+14 0 67,600
nC4 H9 OH ¡ H2 þ C3 H7 CHO 5.0  10+13 0 69,500
R þ nC4 H9 OH ! RH þ nC4 H8 OHa 3HSecondarya
R þ nC4 H9 OH ! RH þ nC4 H8 OHb 2HSecondarya
R þ nC4 H9 OH ! RH þ nC4 H8 OHc 2HSecondarya
R þ nC4 H9 OH ! RH þ nC4 H8 OHd 3HPrimarya
R þ nC4 H9 OH ! RH þ nC4 H9 O 1HPrimarya
nC4 H8 OHa ¡ C2 H5 þ CH3 CHO 3.0  10+13 0 31,000
nC4 H8 OHa ¡ H þ C3 H7 CHO 3.0  10+13 0 37,000
nC4 H8 OHb ¡  OH þ nC4 H8 3.0  10+13 0 33,000
nC4 H8 OHb ¡ CH3 þ C2 H5 CHO 3.0  10+13 0 34,000
nC4 H8 OHb ¡ H þ C3 H7 CHO 3.0  10+13 0 37,000
nC4 H8 OHc ¡ H þ C4 H7 OH 3.0  10+13 0 37,000
nC4 H8 OHc ¡  CH2 OH þ C3 H6 3.0  10+13 0 30,000
nC4 H8 OHd ¡  H þ C4 H7 OH 3.0  10+13 0 36,000
nC4 H8 OHd ¡  CH2 CH2 OH þ C2 H4 3.0  10+13 0 30,000
nC4 H9 O ¡ H þ C3 H7 CHO 3.0  10+13 0 26,000
nC4 H9 O ¡ 1  C3 H7 þ CH2 O 3.0  10+13 0 15,000
nC4 H8 OHd ¡ nC4 H9 O 1.6  10+10 0 16,000
nC4 H8 OHd ¡ nC4 H8 OHa 2.0  10+11 0 19,600
nC4 H8 OHc ¡ nC4 H9 O 1.0  10+11 0 23,000
O2 þ nC4 H8 OHa ¡ HO2 þ C3 H7 CHO 1.5  10+9 0 3500
O2 þ nC4 H8 OHb ¡ HO2 þ 0:67C3 H7 CHO þ 0:33C4 H7 OH 1.5  10+9 0 3500
O2 þ nC4 H8 OHc ¡ HO2 þ C4 H7 OH 1.5  10+9 0 3500
O2 þ nC4 H8 OHd ¡ HO2 þ C4 H7 OH 1.5  10+9 0 3500
O2 þ nC4 H9 O ¡ HO2 þ C3 H7 CHO 1.5  10+9 0 3500
sec-Butanol
sC4 H9 OH ¡ C2 H5 þ CH2 CH OH 2.0  10+16 0 86,500
sC4 H9 OH ¡ CH3 þ  CHOHCH2 CH3 2.0  10+16 0 87,000
sC4 H9 OH ¡ CH3 þ  CH2 CH2 CH2 OH 2.0  10+16 0 87,000
sC4 H9 OH ¡  OH þ  CH2 CH2 CH2 CH3 1.5  10+16 0 93,200
sC4 H8 OHa þ H ¡ sC4 H9 OH 5.0  10+10 0 0

(continued on next page)

2144 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154

Table 1 (continued)

Reactions A n E
 +10
sC4 H8 OHbs þ H ¡ sC4 H9 OH 5.0  10 0 0
sC4 H8 OHbp þ H ¡ sC4 H9 OH 5.0  10+10 0 0
sC4 H8 OHc þ H ¡ sC4 H9 OH 5.0  10+10 0 0
sC4 H9 O þ H ¡ sC4 H9 OH 5.0  10+10 0 0
sC4 H9 OH ¡ H2 O þ 1C4 H8 1.5  10+14 0 67,100
sC4 H9 OH ¡ H2 O þ 2C4 H8 1.0  10+14 0 66,100
sC4 H9 OH ¡ H2 þ CH3 COC2 H5 5.0  10+13 0 69,500
R þ sC4 H9 OH ! RH þ sC4 H8 OHa 1.5HTertiarya
R þ sC4 H9 OH ! RH þ sC4 H8 OHbp 3HPrimarya
R þ sC4 H9 OH ! RH þ sC4 H8 OHbs 2HSecondarya
R þ sC4 H9 OH ! RH þ sC4 H8 OHc 3HPrimarya
R þ sC4 H9 OH ! RH þ sC4 H9 O 1HPrimarya
sC4 H8 OHa ¡ CH3 þ ðCH3 Þ2 CO 3.0  10+13 0 32,000
sC4 H8 OHa ¡ H þ CH3 COC2 H5 3.0  10+13 0 37,000
sC4 H8 OHbp ¡ C2 H5 þ CH3 CHO 3.0  10+13 0 30,000
sC4 H8 OHbp ¡ H þ CH3 COC2 H5 3.0  10+13 0 37,000
sC4 H8 OHbp ¡  OH þ C4 H8 3.0  10+13 0 35,000
sC4 H8 OHbs ¡ CH3 þ C3 H7 CHO 3.0  10+13 0 32,000
sC4 H8 OHbs ¡ H þ CH3 COC2 H5 3.0  10+13 0 37,000
sC4 H8 OHbs ¡ H þ C4 H7 OH 3.0  10+13 0 37,000
sC4 H8 OHbs ¡  OH þ C4 H8 3.0  10+13 0 26,000
sC4 H8 OHc ¡ H þ C4 H7 OH 3.0  10+13 0 37,000
sC4 H8 OHc ¡ CH2 CH OH þ C2 H4 3.0  10+13 0 30,000
sC4 H9 O ¡ CH3 þ C3 H7 CHO 3.0  10+13 0 18,000
sC4 H9 O ¡ C2 H5 þ CH3 CHO 3.0  10+13 0 15,000
sC4 H9 O ¡ H þ CH3 COC2 H5 3.0  10+13 0 23,000
sC4 H8 OHc ¡ sC4 H9 O 1.0  10+11 0 21,000
sC4 H8 OHc¡sC4 H8 OHbp 3.0  10+11 0 19,500
O2 þ sC4 H8 OHa ¡ HO2 þ CH3 COC2 H5 1.5  10+9 0 3500
O2 þ sC4 H8 OHbp ¡ HO2 þ CH3 COC2 H5 1.5  10+9 0 3500
O2 þ sC4 H8 OHbs ¡ HO2 þ 0:67CH3 COC2 H5 þ 0:33C4 H7 OH 1.5  10+9 0 3500
O2 þ C4 H8 OHc ¡ HO2 þ C4 H7 OH 1.5  10+9 0 3500
O2 þ sC4 H9 O ¡ HO2 þ CH3 COC2 H5 1.5  10+9 0 3500
iso-Butanol
iC4 H9 OH ¡ iC3 H7 þ  CH2 OH 2.0  10+16 0 85,000
iC4 H9 OH ¡ CH3 þ CH3 CHCH2 OH 2.0  10+16 0 88,000
iC4 H9 OH ¡  OH þ ðCH3 Þ2 CH CH2 1.5  10+16 0 93,200
iC4 H8 OHa þ H ¡ iC4 H9 OH 5.0  10+10 0 0
iC4 H8 OHb þ H ¡ iC4 H9 OH 5.0  10+10 0 0
iC4 H8 OHc þ H ¡ iC4 H9 OH 5.0  10+10 0 0
iC4 H9 O þ H ¡ iC4 H9 OH 5.0  10+10 0 0
iC4 H9 OH ¡ H2 O þ i  C4 H8 5.0  10+13 0 65,600
iC4 H9 OH ¡ H2 þ ðCH3 Þ2 CHCHO 5.0  10+13 0 69,500
R þ iC4 H9 OH ! RH þ iC4 H8 OHa 3HSecondarya
R þ iC4 H9 OH ! RH þ iC4 H8 OHb 1HTertiarya
R þ iC4 H9 OH ! RH þ iC4 H8 OHc 6HPrimarya
R þ iC4 H9 OH ! RH þ iC4 H9 O 1HPrimarya
iC4 H8 OHa ¡ CH3 þ C3 H7 CHO 3.0  10+13 0 32,000
iC4 H8 OHa¡H ðCH3 Þ2 CHCHO 3.0  10+13 0 37,000
iC4 H8 OHb ¡  OH þ i  C4 H8 3.0  10+13 0 36,000
iC4 H8 OHb ¡ H þ C4 H7 OH 3.0  10+13 0 37,000
iC4 H8 OHb ¡ H þ ðCH3 Þ2 CHCHO 3.0  10+13 0 37,000
iC4 H8 OHc ¡ CH3 þ C3 H7 CHO 3.0  10+13 0 32,000
iC4 H8 OHc ¡  CH2 OH þ C3 H6 3.0  10+13 0 32,000
iC4 H8 OHc ¡ H þ C4 H7 OH 3.0  10+13 0 37,000
iC4 H9 O ! CH2 O þ i  C3 H6 3.0  10+13 0 20,000
iC4 H9 O ! H þ ðCH3 Þ2 CHCHO 3.0  10+13 0 26,000
iC4 H8 OHc ¡ iC4 H9 O 1.0  10+11 0 21,500
O2 þ iC4 H8 OHa ¡ HO2 þ ðCH3 Þ2 CHCHO 1.5  10+9 0 3500
O2 þ iC4 H8 OHb ¡ HO2 þ 0:33ðCH3 Þ2 CHCHO þ 0:67C4 H7 OH 1.5  10+9 0 3500
O2 þ iC4 H8 OHc ¡ HO2 þ C4 H7 OH 1.5  10+9 0 3500
O2 þ iC4 H9 O ¡ HO2 þ ðCH3 Þ2 CHCHO 1.5  10+9 0 3500
tert-Butanol
tC4 H9 OH ¡ CH3 þ ðCH3 Þ2 COH 6.0  10+16 0 81,500
tC4 H9 OH ¡  OH þ ðCH3 Þ3 C 1.5  10+16 0 93,200
tC4 H8 OHb þ H ¡ tC4 H9 OH 5.0  10+10 0 0
tC4 H9 O þ H ¡ tC4 H9 OH 5.0  10+10 0 0
tC4 H9 OH ¡ H2 O þ i  C4 H8 4.5  10+14 0 65,100
R þ tC4 H9 OH ! RH þ tC4 H8 OHb 9HPrimarya
R þ tC4 H9 OH ! RH þ tC4 H9 O 1HPrimarya
tC4 H8 OHb ¡ CH3 þ ðCH3 Þ2 CO 3.0  10+13 0 33,000
tC4 H9 Ob ¡  OH þ i  C4 H8 3.0  10+13 0 37,000
tC4 H9 O ¡ CH3 þ ðCH3 Þ2 CO 9.0  10+13 0 18,000
a
Kinetic parameters of H-abstraction reactions are discussed in Section 3.2.
 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154 2145

thermodynamic and transport properties, is available in CHEMKIN (2) in a flow reactor [1] (Fig. 10),
format [48]. (3) in a jet-stirred reactor at 1 atm and 10 atm [21,57] (Figs. 11
and 12),
4. Numerical methods and simulations (4) ignition delay times measured in shock tubes [17,18] (Figs.
13–17),
The kinetic model described in the previous section was vali- (5) new (Figs. 20 and 21) and previous [21] (Fig. 22) measure-
dated by comparing the results of numerical simulations with ments of structures of n-butanol counterflow non-premixed
new and previous experimental data obtained employing shock flames and
tubes, batch reactor, jet-stirred reactor and flow reactor, and coun- (6) new measurements of structure and species profiles of iso-
terflow non-premixed flames. The DSMOKE code was used to inte- butanol counterflow non-premixed flames (Figs. 23 and 24).
grate numerically the system of differential equations that describe
various aspects of combustion in shock tube and the reactors [49]. In Figs. 9–17 and 20–26, the symbols represent experimental
To predict aspects of combustion that include molecular transport data and the lines are results of numerical simulations. The key
and chemical reactions, a one dimensional laminar flame model features of these simulations and comparisons with experimental
was employed. A detailed description of this code is given else- data are described in the following sections.
where [50]. This code includes multicomponent diffusion and ther-
mal diffusion. Discretization of the differential equations is carried
5.1. Pyrolysis experiments in batch reactors and shock tubes
out using conventional finite differencing techniques for non-uni-
form mesh spacing. The numerical problem corresponds to a large
Early pyrolysis and oxidation experimental and modeling stud-
system of differential–algebraic equations (DAE). The specifically
ies were limited to n-butanol and tert-butanol [53–55]. These
conceived methods and solver routines of BzzMathLibrary [51,52]
studies contributed to the discussion on the role of molecular
are used to handle the complexity of this numerical problem.
and radical reactions in tert-butanol and n-butanol decomposition.
The thermal decompositions of tert-butanol was investigated by
5. Model predictions and comparisons with experimental data
Schultz and Kistiakowsky [53], at low pressures in a static reactor,
in the temperature region of 760–830 K. Similar pyrolysis experi-
Validation of the kinetic mechanism was carried out over a wide
ments both with n-butanol and tert-butanol were conducted by
range by comparing predictions with experimental data obtained:
Barnard [54,55]. An overall decomposition rate of k = 1.5 
1012 exp[56,700/(RT)] [s1] was obtained for n-butanol and
(1) in batch reactors [53–55] and shock tubes [56] (Fig. 9),
k = 3.1  1011 exp[54,500/(RT)] [s1] for tert-butanol. These

6 6
1541 K

5 5
H2Ox10 [mol/cm ]
3

H2Ox10 [mol/cm ]

4
3

3
3
8

2
2
1337 K
1
1 1223 K
0
0
0 20 40 60 80 100 0 200 400 600 800 1000
Time (μ s) Time (μs)

Fig. 9. Predicted and experimental concentration profiles from shock tube pyrolysis of 0.7% tert-butanol in Ar at 0.7–1 atm [56].

1600 200
t-C4H9OH CH3COCH3
CH4
CH3COCH3
i-C4H8 C2H6
Mole Fraction (ppm)

Mole Fraction (ppm)

1200 150 C3H6

C2H4

800 100
CH4

C2H6
400 50
C3H6

C2H4
0 0
0 20 40 60 80 100 0 20 40 60 80 100
Time [ms] Time [ms]

Fig. 10. Predicted and experimental concentration profiles from oxidation of 2000 ppm of tert-butanol in a flow reactor at 1 atm, T = 1027 K and U = 0.67 [1]. Model
predictions are shifted by 25 ms.
2146 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154

1E-3 0.01
CO
C4H9OH
CH4
C3H6

Mole Fraction
Mole Fraction

1E-3 CO2
C4H8
1E-4
C3H7CHO

1E-4

1E-5 1E-5
700 800 900 1000 1100 1200 1300 700 800 900 1000 1100 1200 1300
T [K] T [K]

1E-3 0.01
C2H4 H2O

CH3CHO H2

Mole Fraction
Mole Fraction

C2H6 1E-3 CH2O

C2H2
1E-4

1E-4

1E-5 1E-5
700 800 900 1000 1100 1200 1300 700 800 900 1000 1100 1200 1300
T [K] T [K]

Fig. 11. Predicted and experimental concentration profiles from the oxidation of 1000 ppm of n-butanol in a JSR at 1 atm, U = 1 and s = 0.07 s [21].

n C4H9OH 0.01
1E-3 CO
C3H6
CO2
C4H8 1E-3
CH4
Mole Fraction

1E-4
Mole Fraction

C3H7CHO

1E-4

1E-5
1E-5

1E-6 1E-6
800 900 1000 1100 1200 800 900 1000 1100 1200

T [K] T [K]

1E-3 0.01 H2O

C2H4

CH3CHO H2
1E-3
CH2O
Mole Fraction

C2H6
Mole Fraction

1E-4
C2H2
1E-4

1E-5
1E-5

1E-6 1E-6
800 900 1000 1100 1200 800 900 1000 1100 1200
T [K] T [K]

Fig. 12. Predicted and experimental concentration profiles from the oxidation of 1000 ppm of n-butanol in a JSR at 10 atm, U = 1 and s = 0.7 s [57].
 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154 2147

10000
1000 n C4H9OH 1% mol
t C4H9OH 1% mol
P=1 atm
Ignition delay time (μ s)

P=1 atm

Ignition delay time (μ s)

1000

Φ =1
100
Φ =0.5 Φ =1
100
Φ =0.25 Φ =0.5

Φ =0.25

10
0.60 0.65 0.70 0.75 0.80 0.85 10
0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85
1000/T (K) 1000/T (K)
Fig. 13. Predicted and experimental ignition delay times of Ar/O2 mixtures
Fig. 16. Predicted and experimental ignition delay times of Ar/O2 mixtures
containing 1% n-butanol [17].
containing 1% tert-butanol [17].

10000
10000
s C4H9OH 1% mol
n C4H9OH 0.75% mol
P=1 atm
Ignition delay time (μ s)

P=1 atm

Ignition delay time ( μ s)

1000
1000

Φ =2
Φ =1
100 Φ =1
Φ =0.5
100
Φ =0.5
Φ =0.25

10
0.55 0.60 0.65 0.70 0.75 0.80 0.85 10
0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85
1000/T (K) 1000/T (K)

Fig. 14. Predicted and experimental ignition delay times of Ar/O2 mixtures Fig. 17. Predicted and experimental ignition delay times of Ar/O2 mixtures
containing 1% sec-butanol [17]. containing 0.75% n-butanol [18].

10000 Thus, the predicted activation energy should be about 64,000 cal/
i C4H9OH 1% mol mol in line with the molecular reaction and in agreement with
P=1 atm the experimental value k = 4.8  1014 exp[65,500/(RT)] [s1] pro-
Ignition delay time (μ s)

posed by Schultz and Kistiakowsky [53]. Pyrolysis experiments of

1000 tert-butanol at high temperatures (1200–1600 K) were reported
by Choudhury et al. [56]. Fig. 9 shows the profiles of H2O that is
formed from the pyrolysis process at 1223 K, 1337 K and 1541 K.
Φ =1 Dehydration reaction is dominant, at about 90% at 1300 K and
100 about 70% at 1600 K. The unimolecular radical initiation reactions
Φ =0.5
with high activation energy can only prevail at the high flame
Φ =0.25 temperatures.

10
0.55 0.60 0.65 0.70 0.75 0.80 0.85
5.2. Atmospheric pressure flow reactor
1000/T (K)

Fig. 15. Predicted and experimental ignition delay times of Ar/O2 mixtures
containing 1% iso-butanol [17].
activation energies clearly indicate the relevant role of the chain
radical mechanism. In fact, while the molecular dehydration pro-
cess requires around 65,000 cal/mol for both the alcohols, the
apparent activation energy of the pure radical process is expected
to be in the range of around 55,000 cal/mol. Model predictions of
the pyrolysis of n-butanol show that more than 90% of decomposi-
tion occurs through the radical path, confirming the activation en-
ergy observed by Barnard [54]. On the contrary, similar simulations
of tert-butanol pyrolysis indicate the prevailing role of the molec-
ular dehydration while the radical path is limited to less than 25%.
Norton and Dryer [1] show experimental results for oxidation of
tert-butanol in a flow reactor and a complete analysis of the oxida-
tion of lighter alcohols. The oxidation of tert-butanol was con-
ducted at atmospheric pressure in the Princeton flow reactor at
temperatures of 1027 K and residence times of approximately
0.1 s, in lean condition (U = 0.67) and high N2 dilution. Fig. 10 com-
pares predictions of the kinetic model described here with experi-
mental data of tert-butanol. Note that, following usual practice
employed for comparison with flow reactor data [15], predicted
profiles are shifted by 25 ms in order to match the fuel conver-
sion. This approach and similar corrections, due to non-ideal reac-
tant mixing and/or catalytic effects, were discussed and verified
both numerically and experimentally in the literature [58]. The
predictions agree well with experimental data for all species
2148
except methane. Predictions of concentrations of CH4 are slightly
lower than the measurements. At the conditions investigated here,
more than 80% of tert-butanol decomposition is attributed to the
molecular dehydration reaction. The major product is iso-butene
(iso-C4H8). Acetone (C3H6O) is the primary product of the radical
decomposition path. Propene is mainly derived from the H addi-
tion reaction on iso-butene with the successive demethylation of
the primary iso-butyl radical. The present kinetic model is not able
to properly predict the insignificant formation of carbon monoxide
and carbon dioxide. Model predictions of CO are lower than the
measurements by more than a factor of five. An initial presence
of CO2 observed in the experiments could be attributed to the
impurity in the reactants. It is difficult, however, to explain the for-
mation of more than 100 ppm of CO, mainly due to the limited
importance of the radical oxidation path. A delay in CO formation
was also observed previously in an acetone flame [59]. This devia-
tion could suggest that the decomposition of acetone might di-
rectly occur with a ‘‘roaming methyl” with a possible molecular
path [60]. Thus, CH3COCH3  C2H6 + CO. These deviations require
a more complete kinetic analysis and further experimental
confirmation.
1200
1000
mole fraction [ppm]
iC4H7
800
600
C3H5OH
400
C3H5
200
0 differences in combustion characteristics of the isomers. The re-
0 0. 5 1.0
time [ms]
Fig. 18. Predicted values of mole fraction of intermediate species as a function of
time for oxidation of a stoichiometric mixture of tert-butanol and air at 1400 K and
pressure of 1 bar.
Fig. 19. Relative importance of the different reactions paths of n-butanol and tert-butanol decomposition.
R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154
5.3. n-Butanol oxidation in a jet-stirred reactor at 1 atm and 10 atm
Dagaut and coworkers studied the oxidation of n-butanol in a
jet-stirred reactor with different equivalence ratios, both at 1 atm
[21] and 10 atm [57].
Detailed information on the profiles of several intermediates to-
gether with CO, CO2 and H2O were presented. These experimental
data were used to validate a detailed kinetic model for n-butanol
oxidation. Figs. 11 and 12 show comparison of predictions of the
kinetic model developed here with previous experimental data ob-
tained at 1 atm [21] and 10 atm [57], respectively. Similar compar-
isons with the experiments at 10 atm were recently discussed also
by Black et al. [18]. Figs. 11 and 12 show that the predicted reactiv-
ity of the system is somewhat lower at 1 atm, while a better agree-
ment is observed at 10 atm. Predictions of propene and methane
are slightly lower suggesting the possible greater influence of c-
butanol and b-butanol radicals, respectively. Similar deviations,
at 10 atm, were also observed by Black et al. [18]. It is noteworthy
that the C balance on the experimental data seems to indicate a
systematic overprediction of 5–10%, even if the balance is limited
to the reported/monitored species. This overprediction will be fur-
ther increased if the presence of other species, such as enols, ke-
tene and other minor intermediates is taken into account. A
better agreement should be obtained by normalizing and correct-
ing the experimental information for these systematic deviations.
Similar comparisons between experimental data and model pre-
dictions are also obtained by varying the equivalence ratio, both
CO/100
at 1 and 10 atm.
5.4. Ignition delay times in shock tube
H/10
The high temperature autoignition of the four isomers of buta-
nol have been experimentally studied at atmospheric pressure in a
shock tube and a detailed kinetic mechanism for description of the
high-temperature oxidation of butanol isomers was also discussed
[17]. Ignition delay time were obtained by extrapolating the max-
imum slope in OH emission to the zero value and also by measur-
ing sidewall pressure. This study facilitates comparison of the
1.5 2.0
sults show that n-butanol is the most reactive and tert-butanol
the least reactive [17].
The kinetic model developed here is used to predict the ignition
delay times. The very sharp CH maximum is used to evaluate the
ignition delay times in simulations. Figs. 13–16 compare the
 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154 2149

predictions of ignition delay times, s [ls] as a function of 1000/T, reactive mixture 0.0075. Fig. 17 compares predictions of the kinetic
with experimental data for all the four isomers. Here T is the initial model with experimental data. The apparent activation energies
temperature of the reactive mixture. The experiments and numer- predicted are in the order of 45–50 kcal/mol, while those calcu-
ical simulations were performed for values of equivalence ratio, U, lated using the measured data seem systematically lower. Similar
equal to 0.25, 0.5 and 1.0 with the fuel mass fraction in the reactive deviations were also observed in the previous study on combustion
mixture of 0.01. The predictions of the kinetic model agree with of propanol isomers [15]. This fact clearly points to further exper-
experimental data at low temperatures. Larger deviations between imental and kinetic modeling analysis.
predictions and data are mainly observed at temperatures higher tert-Butanol decomposes to iso-butene (iso-C4H8) in very short
than 1700 K. times (0.1 ms). iso-Butene is the main contributor to the overall
Similar ignition delay times, limited to n-butanol oxidation, ignition delay time. In fact, the important H-abstraction reactions
were also presented by Black et al. [18], with a detailed kinetic dis- on iso-C4H8 forms the resonantly stabilized methyl-allyl radical
cussion. The data was obtained at atmospheric pressure and equiv- (iso-C4H7). The major decomposition reaction of iso-C4H7 leads to
alence ratios of 0.5, 1 and 2 with the fuel mass fraction in the formation of propadiene (C3H4) and methyl radical (CH3). Thus,

1800
Uncoated
0.20
1600 Coated
n C4H9OH
Model
1400
Temperature (K)

Mole Fraction
0.15 O2
1200
CO2
1000
0.10
800 H2O

600 CO
0.05
400
200
0.00
0 1 2 3 4 5 6 7 8 9 10 2 3 4 5 6 7 8
Exit from fuel duct (mm) Exit from fuel duct (mm)

Fig. 20. Profile of temperature, and profiles of mole fraction of reactants, final products, and carbon monoxide as a function of distance from the fuel boundary for non-
premixed n-butanol flame. The strain rate a2 = 100 s1, and the stoichiometric mixture fraction Zst = 0.23. The figure shows temperature profiles, measured using bare
thermocouple and coated thermocouple, and corrected to account for radiative heat losses from the wire.

0.016
H2
0.005 C2H2
C2H4
0.012 0.004 CH4
Mole Fraction

Mole Fraction

C2H6
0.003
0.008
0.002
0.004
0.001

0.000 0.000

2 3 4 5 6 7 8 2 3 4 5 6 7 8
Exit from fuel duct (mm) Exit from fuel duct (mm)

-4
0.0030 8.0x10
C3H6 C4H6
0.0025
n C4H8 -4 p C3H4
6.0x10
Mole Fraction

0.0020
Mole Fraction

CH3CHO C3H8

-4
0.0015 CH2O 4.0x10 a C3H4

0.0010
-4
2.0x10
0.0005

0.0000 0.0
2 3 4 5 6 7 2 3 4 5 6 7
Exit from fuel duct (mm) Exit from fuel duct (mm)

Fig. 21. Profile mole fraction of various species as a function of distance from the fuel boundary for non-premixed n-butanol flame at a value of the strain rate a2 = 100 s1,
and the stoichiometric mixture fraction Zst = 0.23.
2150 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154

iso-C4H7  CH3 + C3H4, with a rate constant of k = 1.0  1013
exp[57,000/(RT)] [s1]. The addition of H to iso-C4H8 forms iso-
butyl radical (iso-C4H9) and tert-butyl radical (tert-C4H9). The
demethylation of iso-C4H9 radicals forms propene (C3H6), a precur-
sor of the stable allyl radical (C3H5). Fig. 18 shows this chemical
evolution. Recombination of OH and allyl radicals explains the for-
mation of propenol (C3H5OH). The main decomposition products of
propenol are acrolein (C3H4O) and H radicals. These species make a
significant contribution to the ignition of the system. The sharp
peak of H radicals corresponds to the maximum of CO concentra-
tion and the ignition of the system with CO2 formation. These ki-
netic considerations clarify that the large deviations observed at
high temperatures in Figs. 13–16 could primarily attributed to
the successive reactions of allyl and methyl-allyl radicals more
than to the primary reactions of butanols. Referring to the compar-
isons reported in Fig. 10, it is evident the relevant role of the
molecular dehydration reaction with formation of iso-C4H8 over
the radical path with acetone formation in tert-butanol oxidation
at 1050 K. Fig. 19 compares the relative importance of the different
reactions paths when n-butanol or tert-butanol are oxidized in air
at 1 bar and at stoichiometric conditions. These values are evalu-
ated at 50% of fuel conversion. At 1600 K, this figure indicates that
for n-butanol the relative importance of chain propagations reac-
tions, molecular reactions and initiation reactions is 50%, 20%
and 30%, respectively.
At low temperatures, in the range 900–1000 K, radical chain
propagation reactions prevail. The importance of the four centre
molecular reactions appears in a different way for the two isomers
tert-butanol and n-butanol. As already observed, molecular dehy-
dration reaction largely dominates the tert-butanol decomposition
in the temperature range 1000–1600 K, while it accounts only for
15–20% of n-butanol decomposition. Unimolecular decomposition
reactions prevail for both of these isomers of butanol at high tem-
peratures, due to the very high activation energies of these radical

2500
0.14
n C4H9OH
2000 0.12
CO
Temperature (K)

Mole Fraction 0.10

1500 CO2
0.08

1000 0.06

0.04
500
0.02

0.00
0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20
Exit from fuel duct (mm) Exit from fuel duct (mm)

12000 2000
C2H4 C2H6

10000 C2H2 n C4H8

1600
Mole Fraction (ppm)

Mole Fraction (ppm)

CH4 C3H6
8000
1200
6000
800
4000

400
2000

0 0
0 2 4 6 8 10 12 0 2 4 6 8 10 12
Exit from fuel duct (mm) Exit from fuel duct (mm)

1400 CH2O C3H8

200
1200 CH3CHO p C3H4
Mole Fraction (ppm)
Mole Fraction (ppm)

1000 150
800

600 100

400
50
200

0 0
0 2 4 6 8 10 12 0 2 4 6 8 10 12
Exit from fuel duct (mm) Exit from fuel duct (mm)

Fig. 22. Profiles of temperature and mole fraction of various species as a function of distance from the fuel boundary for non-premixed n-butanol flame [21] at a value of the
strain rate a2 = 20 s1, and the stoichiometric mixture fraction Zst = 0.5283.
 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154 2151

1800 0.25
1600
0.20
1400 i C4H9OH
Temperature (K)

Mole Fraction
1200
0.15 O2
1000
CO2
800 0.10
H2O
600
400 0.05 CO

200
0.00
0 1 2 3 4 5 6 7 8 9 10 2 3 4 5 6 7 8
Exit from fuel duct (mm) Exit from fuel duct (mm)

Fig. 23. Profile of temperature, and profiles of mole fraction of reactants, final products, and carbon monoxide as a function of distance from the fuel boundary for non-
premixed iso-butanol flame. The strain rate a2 = 100 s1, and the stoichiometric mixture fraction Zst = 0.23. The figure shows temperature profiles, measured using the coated
thermocouple, and corrected to account for radiative heat losses from the wire.

0.015
CH2O
C3H6 0.0024
i C4H8
C2H4
Mole Fraction
Mole Fraction

0.010 C2H6
H2 0.0016
C2H2

0.005
0.0008

0.000 0.0000
2 3 4 5 6 7 2 3 4 5 6 7
Exit from fuel duct (mm) Exit from fuel duct (mm)

-4
5.0x10

-4
4.0x10
CH3CHO
Mole Fraction

-4 a C3H4
3.0x10
C4H6
-4
2.0x10
p C3H4

-4
1.0x10

0.0
2 3 4 5 6 7
Exit from fuel duct (mm)

Fig. 24. Profile mole fraction of various species as a function of distance from the fuel boundary for non-premixed iso-butanol flame at a value of the strain rate a2 = 100 s1,
and the stoichiometric mixture fraction Zst = 0.23.

initiation reactions. Similarly, molecular dehydration of iso-buta- recorded by the uncoated thermocouple is about 200 K higher than
nol accounts for 20%, while the reactions to form both 1-C4H8 that measured using the coated thermocouple. This indicates that
and 2-C4H8 from sec-butanol account for 40%. the chemical reactions are not in equilibrium near the location
where the temperature is the highest. As a consequence, there is
5.5. Non-premixed flame of n-butanol and iso-butanol catalytic heating on the surface of the uncoated thermocouple.
The predicted temperature profile is slightly narrow when com-
Experimental measurements of the structure of counterflow pared with experimental data obtained using the coated thermo-
non-premixed flame of n-butanol, discussed in Section 2, are com- couple. The agreement is considered to be satisfactory. Figs. 20
pared with model predictions in Figs. 20 and 21. Fig. 20 shows the and 21 compare the predicted profiles of various species with
temperature profile measured using the bare (uncoated) thermo- experimental data. The predicted profiles of the reactants n-
couple and the coated thermocouple. The measured values of the C4H9OH, and O2, final products CO2 and H2O, and the key interme-
temperature are the same everywhere except in the vicinity of diates H2 and CO agree well with experimental data. The predicted
the position of maximum temperature. The maximum temperature and measured profiles show that the peak values of CO and H2 are
2152 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154

located at the position where the concentration of fuel is very
small. The peak values of CO2, H2O and temperature are observed
on the right side of the position where the concentration of fuel
is very small. This is consistent with the asymptotic description
of flame structure where the reaction zone is separated into two
layers—an inner layer and an outer layer [3–5]. In the inner-layer,
butanol reacts with radicals and the final intermediates formed are
CO and H2. These intermediates are oxidized to CO2 and H2O in the
outer-layer. The predicted profiles of CH4, C2H2, C2H4, C2H6, C3H6,
C4H6 and C4H8 agree well with the measurements. Butanal was
also detected in the n-butanol flame. The measured concentrations
were very low. Therefore, they are not shown in Fig. 21.
Fig. 22 compares predictions of structure of n-butanol flame
with experimental data of Sarathy et al. [21]. Comparison of the
Figs. 20 and 21 with Fig. 22 highlights the influence of strain rate
and stoichiometric mixture fraction on flame structure. These data
are useful both in order to confirm the experimental measure-
ments and to verify the possible systematic deviations between
model predictions and experiments. There are a number of similar-
ities between the two flames. The overall agreement between the
predicted and measured profiles of intermediate species is satisfac-
tory. Figs. 21 and 22 show that the predicted mole fraction of form-
aldehyde (CH2O) is higher than the measurements. The predicted
mole fraction of acetaldehyde (CH3CHO) in Fig. 21 is less than
the measurements. These deviations are not observed in the
n-butanol oxidation in the JSR experiments reported in Figs. 11
and 12. Although, predictions of butadiene (C4H6) agrees reason-
ably well with the measurements in Fig. 22, they are higher than
the measurements shown in Fig. 21. The same deviations were also
observed between predictions and measurements in Sarathy et al.
[21].
Figs. 23 and 24 compare predictions of the kinetic model of the
structure of non-premixed iso-butanol flame with experimental
data. The iso-butanol flame structure was measured at conditions
identical to those shown in Figs. 20 and 21 for n-butanol. The max-
imum value of the temperature recorded by the uncoated thermo-
couple is about 200 K higher than that measured using the coated
thermocouple. The predicted temperature profile agrees well with
the measurements obtained using the coated thermocouple. The
predicted profiles of the n-C4H9OH, O2, CO2, H2O, CO, H2, C2H4,
C2H2 and C3H6 agree well with the measurements. The asymptotic
flame structure constructed from the profiles of the reactants,
products and major intermediates are similar to that of n-butanol.
Temperature profiles of the n-butanol and iso-butanol flames
shown in Figs. 20 and 23 are similar, due to the same boundary
conditions employed. Comparison of Figs. 21 and 24 show that
the mole fractions of CH2O are similar, while the mole fraction of
CH3CHO is lower in the iso-butanol flame. Trace amounts of but-
anal was observed in the n-butanol flame but none in the iso-buta-
nol flame.
6. Summary and conclusions
A kinetic mechanism, that describes the primary reactions of
pyrolysis and combustion of butanol isomers, is developed. This
mechanism is appended to a previously developed detailed scheme
of pyrolysis and oxidation of hydrocarbon fuels. Employing a sys-
tematic approach for the different reaction classes, including

60 60
n C4H9OH
Laminar Flame Speed (cm/s)
Laminar Flame Speed (cm/s)

n-C3H7OH
50 50

40 40
i-C3H7OH
30 30

20 20

10 10 Veloo et al. 2010

Model
0 0
0.6 0.8 1.0 1.2 1.4 1.6 0.6 0.8 1.0 1.2 1.4 1.6
Equivalence Ratio Φ Equivalence Ratio Φ

Fig. 25. The flame velocity of n-propanol, iso-propanol and n-butanol in air [63,64] as a function of equivalence ratio (P = 1 atm and T0 = 343 K).

1800 1800
n C3H7OH Uncoated Uncoated
i C3H7OH
1600 Coated Coated
1600
Model Model
1400 1400
Temperature (K)
Temperature (K)

1200 1200
1000 1000
800
800
600
600
400
400
200
200
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
Exit from fuel duct (mm) Exit from fuel duct (mm)

Fig. 26. Profile of temperature as a function of distance from the fuel boundary for non-premixed n-propanol flame and non-premixed iso-propanol flame. The strain rate
a2 = 97.5 s1, and the stoichiometric mixture fraction Zst = 0.2449. The figure shows temperature profiles measured using bare thermocouple and coated thermocouple. The
profiles corrected for radiative heat losses.
 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154
four-center molecular dehydration reactions to form butenes, a References
small subset of new primary reactions were proposed for describ-
ing combustion of butyl alcohols. This kinetic model does not in- [1]
[2]
clude low temperature reactions. To our knowledge, the batch [3]
experiments of Cullis and Warwicker [61] with the four isomers [4]
of butanol at 550–700 K are the only low temperature experimen-
tal data. Further low temperature data refer to n-heptane/n-buta- [5]
nol mixtures in the Orleans JSR [62] where the prevailing role of [6]
peroxy radicals from n-heptane covers the n-butanol effect. The [7]
[8]
expected weak effect of low temperature mechanisms of butanol
[9]
isomers requires further dedicated experimental investigation. [10]
New experimental data on counterflow non-premixed flame of [11]
n-butanol and iso-butanol were obtained for mechanism valida-
[12]
tion. Flame structures were measured under the same conditions [13]
for both fuels to elucidate the similarities and differences in com-
bustion characteristics of the two isomers. The measured species [14]
[15]
profiles include those of formaldehyde and acetaldehyde. The val-
idation of the kinetic model was extended to different sets of [16]
experimental data obtained by other investigators in very different [17]
operative conditions and experimental configurations. The agree- [18]
ment between the kinetic model and experimental data was gener-
ally satisfactory, in terms of reactivity and selectivity in major [19]
[20]
products and minor species. The flame structures and overall com-
bustion characteristics of the four butanol isomers are found to be [21]
similar and also similar to those of the two propanol isomers. Sim-
[22]
ilarity of flame velocity in combustible mixtures containing n-pro-
panol and n-butanol further confirms this. Fig. 25 shows that [23]
predictions of the flame velocity of the kinetic model agree well
with experimental data in Ref. [63] and Ref. [64]. The kinetic model [24]
[25]
developed here, together with the reference kinetic parameters of
reactions involving the presence of hydroxyl group, is a useful [26]
starting point for extension to higher alcohols. [27]
[28]
[29]
Acknowledgments
[30]
The Authors acknowledge the valuable discussions with Profes- [31]
sor M. Dente and the useful work of A. Cuoci. The Authors further
[32]
acknowledge Professors F.N. Egolfopoulos, Dr. C.K. Westbrook and
Dr. P.S. Veloo for providing the experimental data on flame velocity [33]
of Fig. 25 prior to their publication. The research at Politecnico di
[34]
Milano is supported by CNR-MAP (biofuel project) and by the Euro-
[35]
pean Union Seventh Framework Programme (FP7/2007-2013) un- [36]
der the DREAM Project Grant Agreement No. 211861. The
research at the University of California at San Diego is supported [37]
by UC Discovery/Westbiofuels, Grant # GCP06-10228.
[38]
[39]
Appendix A. Temperature profiles of n-propanol and iso-
propanol flames [40]
[41]
[42]
An experimental and kinetic modeling study was carried out
previously on n-propanol and iso-propanol combustion [15]. In [43]
this study the structures of counterflow flames of n-propanol and
[44]
iso-propanol were measured using the burner and procedures de-
scribed here. The measurements were made for YF,1 = 0.3, [45]
T1 = 353 K, YO2,2 = 0.233, T2 = 298 K, a2 = 97.5 s1, V1 = 0.235 m/s, [46]
V2 = 0.25 m/s, and L = 10 mm. At these conditions the stoichiome- [47]
tric mixture fraction Zst = 0.2449. In this previous study the mea-
sured temperature profile was not shown, because it was not [48]
[49]
available. The temperature profiles were measured using the bare
and coated thermocouple and are shown in Fig. 26. The procedure [50]
is the same as described here. The temperature measured by the
bare thermocouple is higher than that recorded by the coated ther- [51]
mocouple. This clearly indicates that there is catalytic heating of
the bare wire. Fig. 26 shows that the predictions of the kinetic [52]
model agree with the experimental data.
2153
T.S. Norton, F.L. Dryer, Proc. Combust. Inst. 23 (1990) 179–185.
T.J. Held, F.L. Dryer, Proc. Combust. Inst. 25 (1994) 901–908.
B. Yang, K. Seshadri, Combust. Sci. Technol. 97 (1994) 193–218.
K. Seshadri, F.A. Williams, Reduced chemical systems and their application in
turbulent combustion, in: P.A. Libby, F.A. Williams (Eds.), Turbulent Reacting
Flows, Academic Press, San Diego, CA, 1994, pp. 153–210.
K. Seshadri, Proc. Combust. Inst. 26 (1996) 831–846.
S.C. Li, F.A. Williams, Proc. Combust. Inst. 26 (1996) 1017–1024.
F.N. Egolfopoulos, D.X. Lu, C.K. Law, Proc. Combust. Inst. 24 (1992) 833–841.
N. Marinov, Int. J. Chem. Kinet. 31 (1999) 183–220.
W. Tsang, Int. J. Chem. Kinet. 36 (8) (2004) 456–465.
P. Saxena, F.A. Williams, Proc. Combust. Inst. 31 (2007) 1149–1156.
R. Seiser, S. Humer, K. Seshadri, E. Pucher, Proc. Combust. Inst. 31 (2007) 1173–
1180.
P. Dagaut, C. Togbé, Energy Fuels 22 (2008) 3499–3505.
M. Abián, C. Esarte, A.A. Millera, R. Bilbao, M.U. Alzueta, Energy Fuels 22 (2008)
3814–3823.
A. Frassoldati, A. Cuoci, T. Faravelli, E. Ranzi, Combust. Sci. Technol., in press.
A. Frassoldati, A. Cuoci, T. Faravelli, U. Niemann, E. Ranzi, R. Seiser, K. Seshadri,
Combust. Flame 157 (2010) 2–16.
P. Dagaut, C. Togbé, Fuel 87 (2008) 3313–3321.
J. Moss, A.M. Berkowitz, M.A. Oehlschlaeger, J. Biet, A. Warth, P.A. Glaude, F.
Battin-Leclerc, J. Phys. Chem. 112 (2008) 10843–10855.
G. Black, H. Curran, S. Pichon, J. Simmie, V. Zhukov, Combust. Flame 157 (2010)
363–373.
C. McEnally, L.D. Pfefferle, Proc. Combust. Inst. 30 (2005) 1363–1370.
B. Yang, P. Oßwald, Y. Li, J. Wang, L. Wei, Z. Tian, F. Qi, K. Kohe-Höinghaus,
Combust. Flame 148 (2007) 198–209.
S.M. Sarathy, M.J. Thomson, C. Togbé, P. Dagaut, F. Halter, C. Mountaim-
Rousselle, Combust. Flame 156 (2009) 852–864.
F.A. Williams, Combustion Theory, second ed., Addison-Wesley Publishing
Company, Redwood City, CA, 1985.
A. Liñán, F.A. Williams, Fundamental aspects of combustion, Oxford
Engineering Science Series, vol. 34, Oxford University Press, New York, 1993.
N. Peters, Combust. Sci. Technol. 30 (1983) 1–17.
N. Peters, Turbulent Combustion, Cambridge University Press, Cambridge,
England, 2000.
K. Seshadri, F.A. Williams, Int. J. Heat Mass Transfer 21 (2) (1978) 251–253.
J.H. Kent, Combust. Flame 14 (1970) 279–282.
R.C. Peterson, N.M. Laurendeau, Combust. Flame 60 (1985) 279–284.
E. Ranzi, M. Dente, T. Faravelli, G. Pennati, Combust. Sci. Technol. 95 (1994)
1–50.
A. Galano, J.R. Alvarez-Idaboy, G. Bravo-Perez, M.E. Ruiz-Santoyo, Phys. Chem.
Chem. Phys. 4 (2002) 4648–4662.
R. Atkinson, D.L. Baulch, R.A. Cox, R.F.H.J.A. Kerr Jr., M.J. Rossi, J. Troe, J. Phys.
Chem. Ref. Data 26 (1997) 521–1011.
R. Atkinson, D.L. Baulch, R.A. Cox, R.F.H.J.A. Kerr Jr., M.J. Rossi, J. Troe, J. Phys.
Chem. Ref. Data 28 (1999) 191–394.
E. Ranzi, A. Sogaro, P. Gaffuri, G. Pennati, C.K. Westbrook, W.J. Pitz, Combust.
Flame 99 (1994) 201–211.
H.J. Curran, Int. J. Chem. Kinet. 38 (4) (2006) 250–275.
S.W. Benson, Thermochemical Kinetics, second ed., Wiley, New York, 1996.
H.J. Curran, P. Gaffuri, W.J. Pitz, C.K. Westbrook, A comprehensive modeling
study of n-heptane oxidation, Combust. Flame 114 (1998) 149–177.
M. Dente, G. Bozzano, T. Faravelli, A. Marongiu, S. Pierucci, E. Ranzi, Modeling
of pyrolysis processes in gas and condensed phase, in: G. Marin (Ed.), Advances
in Chemical Engineering, vol. 32, 2007, pp. 52–166.
J. Moc, J.M. Simmie, H.J. Curran, J. Mol. Struct. 928 (2009) 149–157.
C.G. Newman, H.E. O’Neal, M.A. Ring, F. Leska, N. Shipley, Int. J. Chem. Kinet. 11
(1979) 1167–1182.
B.H. Bui, R.S. Zhu, M.C. Lin, J. Chem. Phys. 117 (24) (2002) 11188.
M. Dente, Personal Communication, 2009.
E. Ranzi, A. Sogaro, P. Gaffuri, G. Pennati, T. Faravelli, Combust. Sci. Technol. 96
(1994) 279–325.
E. Ranzi, M. Dente, A. Goldaniga, G. Bozzano, T. Faravelli, Prog. Energy
Combust. Sci. 27 (2001) 99–139.
E. Ranzi, T. Faravelli, A. Frassoldati, S. Granata, Ind. Eng. Chem. 44 (2005) 5170–
5183.
E. Ranzi, Energy Fuels 20 (2006) 1024–1032.
R.J. Kee, F. Rupley, J.A. Miller, The Chemkin Thermodynamic Data Base, Report
SAND86-8215B, Sandia National Laboratories, Livermore, CA, 1987.
R.J. Kee, F. Rupley, J.A. Miller, The Chemkin Thermodynamic Data Base,
Technical Report, Sandia National Laboratories, Livermore, CA, 1989.
http://www.chem.polimi.it/CRECKModeling.
D. Manca, G. Buzzi-Ferraris, T. Faravelli, E. Ranzi, Combust. Theory Modell. 5
(2) (2001) 185–199.
A. Cuoci, A. Frassoldati, T. Faravelli, E. Ranzi, Frequency response of counter
flow diffusion flames to strain rate harmonic oscillations, Combust. Sci.
Technol. 180 (2008) 767–784.
G. Buzzi-Ferraris, Building Numerical Libraries the Object-Oriented Way,
Addison Wesley/Longman, 1993.
G. Buzzi-Ferraris, Numerical Libraries in C++, Politecnico di Milano. <http://
www.chem.polimi.it/homes/gbuzzi>.
2154 R. Grana et al. / Combustion and Flame 157 (2010) 2137–2154
[53] R. Schultz, G.B. Kistiakowsky, J. Am. Chem. Soc. 56 (1934) 395–398.
[54] J.A. Barnard, Trans. Faraday Soc. 53 (1957) 1423–1430.
[55] J.A. Barnard, The pyrolysis of tert-butanol, Trans. Faraday Soc. 54 (1958) 947–951.
[56] T.K. Choudhury, M.C.L.M.C.C.-Y. Lin, W.A. Sanders, Combust. Sci. Technol. 71
(1990) 219–232.
[57] P. Dagaut, S.M. Sarathy, M.J. Thomson, Proc. Combust. Inst. 32 (2009) 229–234.
[58] Z. Zhao, M. Chaos, A. Kazakov, F.L. Dryer, Int. J. Chem. Kinet. 40 (2008) 1–18.
[59] S. Pichon, G. Black, N. Chaumeix, M. Yahyaoui, J.M. Simmie, H.J. Curran, R.
Donohue, Combust. Flame 156 (2009) 494–504.
[60] S. Saxena, J.H. Kiefer, S.J. Klippenstein, Proc. Combust. Inst. 32 (2009) 123–
130.
[61] C.F. Cullis, E.A. Warwicker, Proc. Roy. Soc. 264 (1318) (1961) 392–407.
[62] P. Dagaut, C. Togbé, Energy Fuels 23 (7) (2009) 3527–3535.
[63] P. Veloo, F. Egolfopoulos, C. Westbrook, Studies of n-propanol/air and
isopropanol/air premixed flames. 2009 Fall Meeting of the Western States
Section of the Combustion Institute, 2009.
[64] P.S. Veloo, Y.L. Wang, F.N. Egolfopoulos, C.K. Westbrook, Combust. Flame 157
(2010) 1989–2004.